Optimizes AI workflows by providing expert guidance and implementation patterns for Google Gemini models.
This skill empowers Claude to assist developers in crafting high-performance prompts specifically for the Google Gemini ecosystem. It covers advanced techniques like system instructions, multimodal input handling, and utilizing ultra-long context windows up to one million tokens. Whether you are building complex agents with ReAct patterns or need structured JSON outputs, this skill provides the necessary templates and best practices to leverage Gemini's unique reasoning and multimodal capabilities effectively.
Key Features
01Structured output configuration using JSON Schema and function calling
02Ultra-long context strategies for managing up to 1M+ tokens
03Multimodal prompting patterns for images, audio, and video inputs
04Advanced System Instruction templates for precise behavior control
052 GitHub stars
06Complex reasoning frameworks including Chain-of-Thought and ReAct
Use Cases
01Migrating or optimizing existing prompts for the Gemini 3 and 2.5 model families
02Developing AI agents that require native multimodal reasoning and tool use
03Analyzing massive codebases or document sets using industry-leading context windows
